visual-tokenizer

Tag

Cards List
#visual-tokenizer

RoAR-VAE: Improving Latent Diffusability through Robust Prediction

Hugging Face Daily Papers ↗ · 2026-10-19 Cached

RoAR-VAE提出通过从受损隐变量中鲁棒地预测语义与重建信息来改进视觉tokenizer,在保持重建保真度的同时提升下游扩散模型的可学习性,平均生成基准提升2.19分,小文本重建精度较FLUX.1-VAE提升16.93个百分点。

0 favorites 0 likes
#visual-tokenizer

Unified Multimodal Autoregressive Modeling with Shared Context-Visual Tokenizer is Key to Unification

Hugging Face Daily Papers ↗ · 2026-06-16 Cached

UniAR presents a unified autoregressive framework that uses a single discrete visual tokenizer to bridge visual understanding and generation, achieving state-of-the-art results in image generation and editing.

0 favorites 0 likes
← Back to home

Submit Feedback